Vegetation convergence during early succession on coal wastes: a 6‐year permanent plot study

Question: Does the course of succession on a coal mine restored by hydroseeding converge with a reference community in terms of species composition and vegetation structure? What is the rate of succession on restored areas? How does the balance between local colonization and extinction rates change during succession? Which species group (native or hydroseeded) determines the successional process? Location: Large reclaimed coal mine in the north of Palencia province, northern Spain (42°50′N, 4°38′W). Methods: Between 2004 and 2009 we monitored annually vascular plant species cover in nine permanent plots (20 m² each) at a restored mine; these plots were structured to account for site aspect (north, south and flat). Three identical permanent plots were established in the surrounding reference community and monitored in 2004 and 2009. We used detrended correspondence analysis to assess successional trends and rates of succession, generalized linear mixed models to derive patterns of vegetation structural changes and turnover through time, and Huisman–Olff–Fresco modelling to illustrate response of individual species through time. Results: The three restored mine areas exhibited a successional trend towards the reference community through time, although speed of convergence differed. However, after 6 years the restored sites had diverged considerably and this was greater than the dissimilarity reduction with respect to the reference community. Richness, diversity and native species cover increased linearly through time, whereas hydroseeded species cover decreased. Success of hydroseeded species initially differed in the three areas, and this was negatively related with native species colonization rates. Response patterns through time of ten hydroseeded and 20 most common native species are described. Conclusions: Vegetation structural parameters rapidly converged with the reference community, whereas compositional convergence needed much longer. At the same time, successional composition trajectories and rates were related to site properties (here aspect).     

Topsoil Seed Bank of an Oak–Hickory Forest in Eastern Kentucky as a Restoration Tool on Surface Mines

Typical reclamation practices in the central Appalachian coal region often use compacted spoils as a topsoil replacement, and these soils are revegetated with aggressive grasses and legumes. This restoration approach results in an herbaceous‐dominated landscape with limited natural succession by native flora. An alternative restoration method is to save topsoil prior to mining, stockpile it during mining, and then replace it on uncompacted spoils to “inoculate” the site with native plant species. In an effort to test this approach, vegetation assessments were performed at a relatively undisturbed forested site in Clay County, Kentucky, U.S.A. Eight 15 × 15–m plots were established, and soils from individual plots were used in seed bank studies both in the greenhouse and on loose‐dumped mine spoils. Bulk soil samples were removed from the plots and subjected to cold stratification for 13 weeks, after which seeds were allowed to germinate under greenhouse conditions for 1 year. Additional topsoil (approximately 1.5 m³ from the upper 0–20 cm) was removed from the plots and replaced on fresh spoil in eight 2 × 5–m plots. Controls consisted of uncompacted spoil material substrate only. A total of 105 species emerged in the greenhouse from the seed bank. On the relocated topsoil, 69 species were recorded of which 39 were also observed in pre‐mine vegetation surveys. Ten of the 17 most important pre‐mine forested site species emerged from the relocated topsoil treatments on the mine site. Our results indicate that application of topsoil could enhance plant diversity and native species reestablishment on surface‐mined lands.     

Lack of tree layer control on herb layer characteristics in a subtropical forest,China

Question: Knowledge of the interaction between understorey herb and overstorey tree layer diversity is mostly restricted to temperate forests. How do tree layer diversity and environmental variables affect herb layer attributes in subtropical forests and do these relationships change in the course of succession? Do abundance and diversity of woody saplings within the herb layer shift during succession? Location: Subtropical broad‐leaved forests in southeast China (29°8′18″‐29°17′29″N, 118°2′14″118°11′12″E). Methods: A full inventory of the herb layer including all plants below 1‐m height was done in 27 plots (10 × 10 m) from five successional stages (<20, <40, <60, <80 and ≥80 yr). We quantified the contribution of different life forms (herbaceous, woody and climber species) to herb layer diversity and productivity and analysed effects of environmental variables and tree layer diversity on these attributes. Results: Herb layer composition followed a successional gradient, as revealed by non‐metric multidimensional scaling (NMDS), but diversity was not correlated to the successional gradient. There was no correlation of diversity across layers. Herb layer productivity was neither affected by tree layer diversity nor by herb layer diversity. Although abundance of woody species in the herb layer decreased significantly during succession, woody species contributed extraordinarily to herb layer species diversity in all successional stages. All environmental factors considered had little impact on herb layer attributes. Conclusions: The subtropical forest investigated displays an immense richness of woody species in the herb layer while herbaceous species are less prominent. Species composition of the herb layer shows a clear successional pattern, however, the presence or absence of certain species appears to be random.     

Natural revegetation on forest topsoil and subsoil along roadsides in boreal forest

Questions: Does natural revegetation from indigenous soil improve the restoration success of roadside areas? What are the effects of topsoil, subsoil and fertilization on natural revegetation? Location: Akershus county, SE Norway (10°25′ E, 59° 44′ N). Methods: We used a recently constructed road through a boreal coniferous forest for a three year (2000–2002), fully replicated revegetation experiment (six replications). Treatments were soil type (two levels; one topsoil and one subsoil type) and fertilization (two levels; NPK and unfertilized control). Ordination methods, constrained ordination methods as well as univariate statistical methods, such as Wilcoxon's signed‐rank test and correlation analysis, were used to assess the relative importance and significance of treatments on the plant species composition. Results: There was no fertilization effect on species composition. The species composition on both soil types was stabilised by the second year. The species dominating the topsoil were more in accordance with the indigenous vegetation than was the case on the subsoil. The significant difference in species composition among blocks, persisting for the entire study period, indicated that local factors are important determinants of the outcome of revegetation. Conclusions: Unfertilized topsoil provides a revegetation result in better accordance with the indigenous vegetation than does subsoil.     

Soil moisture dynamics and restoration of self‐sustaining native vegetation ecosystem on an open‐cut coal mine

Post‐mining landscape reconstruction on open‐cut coal mines aims to support restoration of self‐sustaining native vegetation ecosystems that in perpetuity require no extra inputs relative to unmined analogs. Little is known about the soil moisture retention capacity of the limited layer of topsoil replaced (often <30 cm deep), impacts of deep ripping of the profile, and the combined impacts of these on plant available water during the mine restoration process. We examined changes in soil moisture parameters (soil water potential, Ψ, and soil water content, Θ) daily using automated soil sensors installed at 30 and 45–65 cm depths on mine restoration sites aged between 3 and 22 years and on adjacent remnant vegetation sites following heavy rainfall events at Meandu mine, southeast Queensland, Australia. Consistent patterns in soil moisture attributes were observed among rehabilitated sites with generally marked differences from remnant sites. Remnant site soil profiles had generally higher Θ after drying than rehabilitated sites and maintained high Ψ for extended periods after rain events. There was a relatively rapid decline of Ψ on reconstructed soil profiles compared with remnant sites although the times of decline onset varied. This response indicated that vegetation restoration sites released soil moisture more rapidly than remnant sites but the rate of drying decreased with increasing rehabilitation age and increased with increasing tree stem density. The rapid drying of mine rehabilitated sites may threaten the survival of some remnant forest species, limit tree growth, and delay restoration of self‐sustaining native ecosystem.     

Spontaneous vegetation succession in human‐disturbed habitats: A pattern across seres

Abstract. Vegetation samples from 15 successional seres in various disturbed habitats in the western part of the Czech Republic were analysed to detect possible trends. For particular seres, data on species cover were available from the onset to 10–76 yr of succession. All seres started on bare ground. Species which attained at least 1% cover in any sere in any year were used as input data for Canonical Correspondence Analysis, assessing the effect of time as the environmental variable, for Detrended Correspondence Analysis and TWINSPAN classification. Two distinct groups ofseres were distinguished: ‘ruderal’, occurring in agricultural, industrial or urban landscapes altered by men, usually on fertile sites; and ‘non‐ruderul’, occurring in less altered, mostly forested landscapes, usually on acid, nutrient‐poor and wetter soils. The former type of succession starts with ruderal annuals, being followed by ruderal perennials. In the latter case non‐ruderal clonal perennials prevail from the onset of succession. The landscape frame is emphasized, beside site environmental conditions, as influencing the type of succession. The character of species attaining dominance in succession, participation of dominant woody plants and the character of late successional stages, i.e. features important from the point of view of potential restoration of human‐disturbed habitats, are discussed.     

Enhancement of late successional plants on ex-arable land by soil inoculations

Restoration of species-rich grasslands on ex-arable land can help the conservation of biodiversity but faces three big challenges: absence of target plant propagules, high residual soil fertility and restoration of soil communities. Seed additions and top soil removal can solve some of these constraints, but restoring beneficial biotic soil conditions remains a challenge. Here we test the hypotheses that inoculation of soil from late secondary succession grasslands in arable receptor soil enhances performance of late successional plants, especially after top soil removal but pending on the added dose. To test this we grew mixtures of late successional plants in arable top (organic) soil or in underlying mineral soil mixed with donor soil in small or large proportions. Donor soils were collected from different grasslands that had been under restoration for 5 to 41 years, or from semi-natural grassland that has not been used intensively. Donor soil addition, especially when collected from older restoration sites, increased plant community biomass without altering its evenness. In contrast, addition of soil from semi-natural grassland promoted plant community evenness, and hence its diversity, but reduced community biomass. Effects of donor soil additions were stronger in mineral than in organic soil and larger with bigger proportions added. The variation in plant community composition was explained best by the abundances of nematodes, ergosterol concentration and soil pH. We show that in controlled conditions inoculation of soil from secondary succession grassland into ex-arable land can strongly promote target plant species, and that the role of soil biota in promoting target plant species is greatest when added after top soil removal. Together our results point out that transplantation of later secondary succession soil can promote grassland restoration on ex-arable land.     

Two centuries of vegetation succession in an inland sand dune area,central Netherlands

Questions: (a) What are the rates and directions of vegetation succession in an inland sand‐dune system? (b) What are the differences in successional trajectories in different relief types? and (c) Is it possible to preserve the last areas of still active dunes and under what circumstances? Location: The study sites were located in the northern part of the Veluwe Region, central Netherlands; longitude 5°44′ E, latitude 52°20′ N, altitude 9 to 24 m a.s.l. Methods: Vegetation and relief mapping was conducted in three permanent plots, 200 m × 200 m in size, in 1988 and 2003. Phytosociological relevés (2400) were recorded in each 10 m × 10 m subplots. Age of woody species was determined by wood coring. Geographic Information System, ordination analyses, and TWINSPAN were used for data exploration and elaboration. Results: A total of 70 vascular plants and 19 bryophytes were recorded over successional stages spanning approximately190 years. The following dominant species formed the sequence of successional stages, but not all participated in all relief types: Ammophila arenaria, Festuca arenaria, Corynephorus canescens, Festuca ovina and Agrostis capillaris, and pine forest dominated in its herb layer at first by Deschampsia flexuosa and later by either Empetrum nigrum, Vaccinium myrtillus or Vaccinium vitis‐idaea. Conclusions: The successional trajectory is basically unidirectional for more than 100 years; no clear multiple successional pathways were observed, as is frequent in coastal dunes. Successional divergence was observed after approximately 130 years in the composition of the herb layer in the closed pine forest. The obvious vegetation heterogeneity in the still active sand‐blown area is related to differences in timing of vegetation establishment on particular relief types, thus the succession exhibits a terrain‐dependent asynchronous character. We conclude that the last patches of still‐active sand dunes can be preserved only by repeated strong artificial disturbances.     

Ecological succession in areas degraded by bauxite mining indicates successful use of topsoil

Brazilian ironstone outcrops (cangas) are nutrient‐poor stressful habitat dominated by slow‐growing woody species with high biodiversity and unique evolutionary history. Mining has produced great impacts on this ecosystem. Spontaneous regeneration of abandoned canga mined areas has not been observed. One of the active methods most widely used for ecological restoration in environments where soil has been lost or severely degraded is topsoil transposition due to the physical, chemical, and microbiological improvement of the substrate, in addition to the seed bank. Thus, plant succession was monitored for 40 months after topsoil transposition in a canga area degraded by aluminum mining, without any other type of management. A completely randomized design with 70 permanent plots (1 × 1 m) was used. Annual phytosociological surveys were carried out and floristic and vegetational spectra were constructed with the life‐forms proposed by Raunkiaer. Floristic composition was compared with a reference site. Overall, 105 species were identified. Both flora and vegetation changed over time, increasing resemblance to the reference areas. The floristic and vegetational spectra after 4 years of topsoil deposition are similar to pristine ones. The vegetation spectrum showed an increase in the dominance of phanerophytes and hemicryptophytes, while therophytes reduced their proportion. The early successional stage is dominated by weeds, like in other canga restoration studies, but did not impede the native species regeneration. Cangas's species recruited well from transposed topsoil. Unlike other studies with fertilized topsoil, our findings show the efficiency of topsoil transposition to provide initial conditions for the ecological restoration of this ecosystem.     

Can Establishment Characteristics Explain the Poor Colonization Success of Late Successional Grassland Species on Ex‐Arable Land?

Abstract Many grassland restoration projects on former arable land face problems because early successional grassland species establish vigorously and persistently but late successional grassland species fail to establish. Differences in establishment characteristics of early and late successional species might provide an explanation for the failure of many late successional species to colonize grasslands on ex‐arable land. I examined whether early and late successional species had different establishment rates in the initial years of a grassland succession, whether a specific establishment stage (seedling emergence, mortality or growth) could be identified as the key process controlling establishment, and what management would enhance the establishment of late successional grassland species. Seeds of three early and three late successional species were sown separately in ex‐arable plots with bare soil, 1‐year‐old vegetation, and 2‐year‐old vegetation. Emergence, mortality, and seedling growth were monitored for 1 year. Early successional species established successfully in the bare soil plots but failed to establish in plots with 1‐ and 2‐year‐old vegetation. Late successional species showed either lower establishment rates in the younger succession stages or decreased establishment with succession that nevertheless resulted in significant establishment in the oldest plots. Seedling emergence proved to be the key factor determining the establishment pattern of early and late successional species. In absolute numbers, emergence of late successional species was, however, similar or higher than that of early successional species, even in the earliest succession stage. The poor establishment of late successional species on former arable land could therefore not be explained solely by differences in establishment characteristics between early and late successional grassland species. Competitive processes between early and late successional species later in the life cycle probably play an important role. The results do point out that establishment of late successional species can be promoted by creating vegetative cover from the start of the restoration effort.     

Secondary Succession and Indigenous Management in Semideciduous Forest Fallows of the Amazon Basin1

To the discussion on secondary succession in tropical forests, we bring data on three under‐addressed issues: understory as well as overstory changes, continuous as opposed to phase changes, and integration of forest succession with indigenous fallow management and plant uses. Changes in vegetation structure and species composition were analyzed in secondary forests following swidden agriculture in a semideciduous forest of Bolivian lowlands. Twenty‐eight fallows, stratified by four successional stages (early = 1–5 yr, intermediate = 6–10 yr, advanced = 12–20 yr, and older = 22–36 yr), and ten stands of mature forests were sampled. The overstory (plants ≥5 cm diameter at breast height [DBH]) was sampled using a 20 × 50 m plot and the understory (plants <5 cm DBH) in three nested 2 × 5 m subplots. Semistructured interviews provided information on fallow management. Canopy height, basal area, and liana density of the overstory increased with secondary forest age. The early stage had the lowest species density and diversity in the overstory, but the highest diversity in the understory. Species composition and abundance differentiated mature forests and early successional stage from other successional stages; however, species showed individualistic responses across the temporal gradient. A total of 123 of 280 species were useful with edible, medicinal, and construction plants being the most abundant for both over‐ and understories. Most of Los Gwarayo preferred mature forests for making new swidden, while fallows were valuable for crops, useful species, and regenerating timber species.     

Soil seed bank and driftline composition along a successional gradient on a temperate salt marsh

Abstract. This study focuses on the relationship between vegetation succession and soil seed bank composition on the Schiermonnikoog (The Netherlands) salt marsh over 100 yr. The importance of driftline material in seed dispersal and the relationship with succession is also investigated. The results indicate that the majority of species have a transient or short‐term seed persistent bank. Seeds of most species are able to float over the salt marsh and become concentrated in the driftline higher up the marsh. After plants have established a seed bank forms, which disappears when vegetation is replaced by later‐successional species. Exceptions are Spergularia mar‐itima, which is still present in the seed bank of late successional stages, and Juncus gerardi and Glaux maritima, which appear in the seed bank of early successional stages, but are absent in the vegetation. Based on the results of this study constraints and possibilities for salt‐marsh restoration by de‐embankment are discussed.     

Identifying Native Vegetation for Reducing Exotic Species during the Restoration of Desert Ecosystems

There is currently much interest in restoration ecology in identifying native vegetation that can decrease the invasibility by exotic species of environments undergoing restoration. However, uncertainty remains about restoration's ability to limit exotic species, particularly in deserts where facilitative interactions between plants are prevalent. Using candidate native species for restoration in the Mojave Desert of the southwestern U.S.A., we experimentally assembled a range of plant communities from early successional forbs to late‐successional shrubs and assessed which vegetation types reduced the establishment of the priority invasive annuals Bromus rubens (red brome) and Schismus spp. (Mediterranean grass) in control and N‐enriched soils. Compared to early successional grass and shrub and late‐successional shrub communities, an early forb community best resisted invasion, reducing exotic species biomass by 88% (N added) and 97% (no N added) relative to controls (no native plants). In native species monocultures, Sphaeralcea ambigua (desert globemallow), an early successional forb, was the least invasible, reducing exotic biomass by 91%. However, the least‐invaded vegetation types did not reduce soil N or P relative to other vegetation types nor was native plant cover linked to invasibility, suggesting that other traits influenced native‐exotic species interactions. This study provides experimental field evidence that native vegetation types exist that may reduce exotic grass establishment in the Mojave Desert, and that these candidates for restoration are not necessarily late‐successional communities. More generally, results indicate the importance of careful native species selection when exotic species invasions must be constrained for restoration to be successful.     

Primary succession on lava domes on Terceira (Azores)

Questions: Are the vegetation attributes significantly different among lava domes and among geomorphologic units as a result of age and soil features? Are the successional rates equal in all the geomorphologic units of the domes? Are the colonizer species of lava domes totally replaced by other species in the late successional stages? Location: Terceira Island, Azores (Portugal). Methods: Three comparable domes of 240, 370 and 2080 yr old were selected. Data on floristic composition, vegetation bioarea (area occupied by plant species in transects), structure, demography and soil nutrients were collected. Quantitative and qualitative changes along the succession gradient were also analyzed. Results: Vegetation attributes were consistent with the successional stage of each dome in the primary sere; however, the geomorphologic units did not follow the same pattern. The influence of the rates of plant colonization and soil formation are responsible for the decrease of the successional rates from footslopes, to summits, to slopes. The vegetation successionally changes from Juniperus scrub, to Juniperus wood and forest, and there is little species replacement since the similarity in species composition is high between the 3 domes. Conclusions: This is a special type of direct succession that takes place mainly through an increase in biomass and structural complexity. We observed small wooded areas in the fissures of very young domes that are samples of later successional forests — a phenomenon that we call ‘zoom effect’.     

Topsoil Stockpiling Versus Exposure to Traffic: A Case Study on In situ Uranium Wellfields

Management of soils on areas subjected to planned disturbance, such as surface mine sites, is crucial to post‐disturbance restoration. We conducted a study to determine which of the following two topsoil management strategies resulted in less of a negative impact to the soil and vegetation resources and better promoted site restoration on in situ uranium wellfields: (1) removal of topsoil from an entire wellfield and stockpiling until mining activity is complete; or (2) leaving the majority of topsoil on the wellfield and allowing it to be exposed to disturbance associated with wellfield development activity (primarily heavy vehicle traffic). Our study compared selected soil properties from areas on in situ uranium wellfields managed by the two strategies stated above and with adjacent, relatively undisturbed sites. Vegetation reestablishment was assessed on sites where topsoil was left in place. Results indicated that levels of vehicular traffic on wellfields did not cause significant soil compaction and that removal and stockpiling of topsoil results in more negative impacts than disturbance inflicted when topsoil is left in place.     

Secondary succession on sandy old‐fields in Hungary

Abstract. Development of semi‐natural vegetation has recently been a primary concern of restoration efforts. A primary management question is whether active intervention is required or spontaneous secondary succession could suffice. We studied 54 old‐fields in central Hungary, which differed in time since abandonment but which had similar environmental conditions and management histories. The sites were grouped into four age groups according to the time elapsed since cultivation abandonment: 1–5, 6–10, 11–23 and 24–33 yr. In each old‐field we recorded the species and estimated their abundances. We grouped species in two ways: according to life form (annuals, biennials, perennials, woody plants) and according to coenological behaviour (weeds, sand and steppe generalists, specialists). We analysed the changes in species number and abundance in these categories as a function of site age. Contrary to other successional studies, the total number of species did not change significantly among the four age groups. A significant change was detected between the first two age groups as to life‐form composition. Species number and abundance of annuals decreased, while the perennials and woody plants increased. As to coenological behaviour, species number changed only in the first two age groups, while abundance changed in the first three. Weeds quickly disappeared and specialists established and spread, while the species number and abundance of generalists did not change significantly. We concluded that the basic shifts in species composition are almost completed within 10 yr. Most of the late successional species colonized and weeds disappeared. We conclude that there was no need for active intervention in this system: the spontaneous secondary succession leads to semi‐natural vegetation.     

Fifty‐seven years of composition change in the eastern boreal forest of Canada

Question: In the boreal forest of eastern Canada, how does forest vegetation change in the sustained absence of fire? Location: Eastern boreal forest in Quebec's North Shore region, Canada (49°30′–50°00′N; 67°30′–68°35′W). Methods: Aerial photos from three different periods (1930, 1965 and 1987) were used to characterize changes in vegetation composition in 23 scenes of 200 ha. Time since fire, presence of secondary disturbances and data on soil and topographic variables were obtained. Ordination and clustering techniques were used to define compositional trajectories of change over the 57‐yr period. These trajectories were further grouped into pathways based on compositional changes, time since fire and preferential deposit‐drainage types. Results: Among the 26 compositional trajectories, three successional pathways were distinguished. Two start post‐fire succession with a dominance of intolerant hardwood. In one of these, this is followed by an increase in Abies balsamea, while in the second the importance of Picea mariana increases with time. In the third pathway P. mariana is an important component from the outset. In this pathway, we observed modest fluctuation in the relative dominance of P. mariana and A. balsamea and variation in stand structure. Conclusion: The boreal forest vegetation of Eastern Canada is diverse and dynamic even in the absence of fire, notably under the influence of partial disturbances. Such disturbances can be associated with changes in composition or stand structure. The development of management strategies aimed at maintaining stand diversity by emulating a broader variety of partial and secondary disturbances should be encouraged.     

Early seral trends in plant community diversity on a recontoured surface mine

The initial floristic composition and structure of vegetation on a recontoured surface mine in southwestern Wyoming was experimentally manipulated to test ideas concerning successional convergence. Experimental treatments involving the planting pattern and density of shrubs and the storage of topsoil were evaluated on the basis of their ability to divert the successional trajectories of the redeveloping vegetation and to increase species diversity. The change in compositional similarity along a topographic gradient was also monitored.The influence of initial composition and soil storage generally declined with time; successional convergence among treatments over the four years of the study was largely monotonic. The planting of shrubs and the storage of topsoil did, however, influence the rate of plant succession. There were no persistent differences in alpha diversity among the experimental treatments. Compositional dissimilarity along the topographic gradient, as expressed by beta diversity, increased with time.     

A test of ecological succession hypotheses using 55‐year time‐series data for 361 boreal forest stands

Aim There has been much work on succession over many decades, but succession fundamentals are still debated because of the reliance on chronosequences and dendrochronological reconstruction, both of which are problematic approaches. Here we use time‐series data to test four hypotheses that lie at the heart of successional theory: (1) the neighbourhood effect hypothesis – tree species abundance is time dependent; (2) the density‐dependence hypothesis – a rare species is more favoured over time; (3) the resource ratio hypothesis – species that can grow at the lowest resource level tend to dominate resource limited sites through succession; and (4) the intermediate disturbance hypothesis – intermediate disturbances increase the abundance of rare species. Location Central boreal forest in Canada (47°50′–50°10′ N; 80°10′–85°50′ W). Methods We used repeated measurements from sequential aerial photography and ground surveys for 361 fire‐origin stands that were measured over a c. 55‐year period. Results Shade‐intolerant Pinus banksiana decreased, tolerant Thuja occidentalis increased, intolerant Populus spp. and Betula papyrifera displayed a U‐shaped trend, and intermediate‐tolerant Picea spp. and tolerant Abies balsamea did not change with time since fire, showing evidence of negative, positive, or neutral neighbourhood effects. Species either persisted for longer or increased more in non‐conspecific stands, and had higher increases in abundance when associated with species of contrasting shade tolerance, supporting the density‐dependence hypothesis and indicating shade‐tolerance complementarity as a mechanism for coexistence. Resource‐poor soils favoured those species capable of tolerating limited resources, whereas rich soils permitted invasion and promoted mixtures, supporting the resource ratio hypothesis. Intermediate disturbances increased the invasibility of rare species. Main conclusions Contrary to previous studies where time since a stand‐replacing disturbance is the sole predictor for succession, our study shows that time alone is either an insufficient predictor or is irrelevant to species dynamics in the boreal forest. Rather, density dependence, site resource and intermediate disturbances are key mechanisms in species dynamics and coexistence over time.